How a Small Site Out-ranks Big Ones: Topical Authority

Atamyrat Hangeldiyev

Systems Architect

January 26, 2026

Topical authority is the recognition a search engine gives a site for covering a defined subject more completely and coherently than anyone else competing for it, so it ranks for that subject without out-spending its rivals on links, in the context of small and mid-sized businesses going up against larger incumbents. It is not link-building and it is not "publish more". It is being demonstrably the most complete and best-connected source on a topic narrow enough that a small business can actually own it. That is the single biggest reason a small site can beat a big one in the answer era, and it is the entire subject of this guide.

A thirty-page niche site passed a national brand on the queries that paid, and the brand had roughly a hundred times its backlinks and four pages on the subject. The niche site sold one category of industrial filtration media. The national brand sold everything in the category and a great deal more, ranked everywhere on broad terms, and had a domain that link tools loved. On "how to choose filtration media for [a specific contaminant]" and the dozen real buyer questions next to it, the small site was first and the brand was on the second results screen or absent. The moment it became clear why was watching a procurement engineer's path: they did not search the brand, they searched their problem, and the small site was the only place that resolved it end to end, from how to read the spec to which substitution was safe to what changes under heat. The brand had the links. The small site had the topic, completely and connectedly, and the engine ranked the source that covered the subject, not the domain that had been pointed at the most. That gap, between having the links and being the source, is what topical authority is, and it is why a small site wins.

This guide explains that mechanism and hands the rest off at the seams. It explains that complete, connected coverage earns authority and why that beats a bigger budget, so a non-technical owner finishes knowing why a small site can outrank a big one and what coverage they would have to build. It does not teach the model that explains why an engine clusters topics and entities the way it does, it does not give the step-by-step procedure to produce a topical map, and it does not teach the internal-link architecture or the page-level snippet craft. Each of those is its own guide, and this page links to them where the boundary falls. The job here is to make the authority mechanism sharp enough to act on.

How a small site outranks a big one: what topical authority actually is

Topical authority is the engine's judgment that your site is the source on a subject, formed from how completely and coherently the site covers that subject, not from how many other sites point at your domain. The keyword era measured a page against a phrase. The link era measured a domain against a link count. Topical authority measures a site against a topic: does this site cover the subject as a connected whole, resolve the real questions inside it, and stay on it consistently enough that a machine can conclude this is what the site is about and is good at. A small business can lose the phrase contest and the link contest and still win the topic, because the topic contest rewards something a focused operator can do that a broad incumbent usually does not.

The reason this works is that a modern engine is trying to find the best source for a need, not the most-pointed-at domain in the vicinity of that need. When it has resolved that a site genuinely and completely covers a defined subject, links from elsewhere become one input among several rather than the input. A complete, coherent treatment of a narrow topic is direct evidence of the thing the engine actually wants. A large backlink number is an indirect, increasingly noisy proxy for it. Given the direct evidence, the engine does not need to lean as hard on the proxy, which is exactly why the thinly-linked specialist now beats the heavily-linked generalist on the specialist's own subject.

Why the engine ranks the most complete source on a topic, not the most-linked domain

A modern engine ranks the source that best covers and resolves a subject because that is what it is built to surface: the page and the site that actually answer the thing a person needs, judged across the whole site rather than by a domain-level link tally. Coverage is causal evidence of being a good source. Link volume is correlational and gameable, and the engine has spent years getting less dependent on it precisely because it was gamed. When a site demonstrably covers a subject completely and connectedly, it has handed the engine the thing the link count was only ever a weak stand-in for.

Take a regional commercial-plumbing company against a national plumbing franchise. The franchise has thousands of backlinks, a domain every link tool rates highly, and a handful of thin pages on commercial backflow testing because it is a small slice of what it does. The regional company has no link budget worth mentioning and fifteen connected pages that resolve commercial backflow end to end: what a test actually checks, why an assembly fails, what a municipality requires and on what cadence, what a failed test costs a building, how a retrofit decision is made honestly. On "commercial backflow testing requirements [region]" and the cluster of real questions a facilities manager has, the regional company outranks the franchise, because it is unmistakably the source on that subject and the franchise merely mentions it. The franchise's link advantage does not transfer, because the engine is not ranking the domain with the most links. It is ranking the source that covers the subject, and on this subject that is the small company.

An example: the thin site that passed a national brand on the queries that paid

The filtration-media site from the opening is the cleanest worked case, so here is the full anatomy of why it won. The national brand's footprint on the subject was four pages: a category landing page heavy on the head term, two thin product pages, and a blog post that defined the term and pivoted to a contact form. Link tools showed the brand's domain towering over the small site. Every conventional signal said the brand should win.

It did not, on the queries that produced orders. The small site had thirty pages that formed a single connected body on filtration media for a defined set of contaminants: how to read a media spec without getting it wrong, the two failure modes that get the wrong media specified, what changes under temperature and flow, how to validate a substitution before it shuts a line down, what each contaminant class actually requires. Each page resolved one real question a buyer has, and the pages linked to each other so the subject was covered as a whole, not as scattered posts. A procurement engineer searching their exact problem landed on the page that resolved it, found the rest of their questions one click away, and never needed the brand. The engine read the same structure the engineer did: this site is the source on this subject. The brand's hundred-to-one link advantage did not move those queries, because none of those links said the brand was the source on filtration media for those contaminants, and thirty connected pages that resolved the subject said exactly that about the small site. The lesson is not that links are worthless. It is that on a defined subject a small business can own, being the complete connected source beats being the better-linked domain, and the small business can build the first thing and cannot out-buy the second.

Coverage and connection beat the backlink pile

Complete, connected coverage of a defined subject beats raw backlink volume for the queries an SMB can realistically win, because those queries are decided by which site is the source on the subject, and coverage is how a site becomes the source while a link pile is not. This is the load-bearing claim of the guide. If it holds, and on the queries a small business can actually win it substantially does, then the budget a small business is told to spend on links is being aimed at the wrong lever, and the work that actually moves these rankings is work a focused operator can do and a generalist usually will not.

The link pile is the wrong lever here for a structural reason, not a moral one. Backlinks were always a proxy for "other people treat this as a credible source". On a broad, generic topic, where the engine has weak direct evidence of who the real source is, that proxy still carries weight. On a defined, specific subject, where a site can directly demonstrate complete coverage, the engine has strong direct evidence and leans on the noisy proxy far less. A small business cannot win the broad generic topic against an incumbent and does not need to. It can win the defined specific subject, and that is precisely the terrain where coverage beats links.

Why raw backlink volume is the wrong lever for the queries you can win

Raw backlink volume is the count of external links pointing at a domain, sold by link brokers as the thing that decides rankings. It is the wrong lever for an SMB because the queries an SMB can win are decided by topical relevance and completeness on a specific subject, and a link count says nothing about whether a site is the complete source on that subject. A thousand links that have nothing to do with filtration media do not make a site the source on filtration media. Thirty pages that resolve filtration media do.

The expensive version of this misunderstanding is the year a regional electrical contractor spent buying links because a vendor said the national competitor outranked it because it had more. The links were bought, the count went up, the rankings on the queries that produced jobs did not move at all. What finally moved them was a dozen genuinely complete, well-connected pages on the specific commercial work the contractor wanted, "service upgrade requirements for [building type]", "what a code-compliance retrofit actually involves", "how to tell a panel needs replacing not repair", each resolving a real question and linked into a coherent whole. The links never made the site the source on that subject. The coverage did, in less time and for less money than the link campaign that produced nothing. The lever was never the count. It was being the source, and the count cannot buy that.

Why a bought domain rating does not move the rankings that matter

A bought domain rating is a third-party vendor score raised by acquiring links in bulk to make a domain look more authoritative in a tool. It does not move the rankings that matter because the engine does not rank that score, and the score is not the thing the engine is measuring. A higher number in a link tool can sit directly next to flat or falling rankings on the queries a business actually needs, because the tool's number and the engine's judgment of who the source is on a subject are two different things, and only one of them decides the result page.

The sharpest version of this is the case where the dashboard and the rankings point in opposite directions at the same time. A B2B equipment distributor's tool-reported domain rating rose substantially over two quarters under a link-buying program while, on the same dashboard, the queries that produced quotes either held flat or slipped, and the competitor taking those queries carried a lower tool number and a deeper, better-connected body of content on the exact equipment categories buyers compared. The dashboard number is a vendor's model of authority; topical authority is the engine's judgment of who the source is. The two can move in opposite directions precisely because they are measuring different things, and only the one the vendor cannot sell you is the one that pays.

The mechanism behind the mechanism: how the engine clusters the topic

Underneath "coverage earns authority" sits a deeper question this guide deliberately does not answer: why a search engine clusters topics, recognizes entities, and decides that a body of pages is "about" a subject at all. Coverage and connection are how a small site earns authority. The model of entities, their attributes, and the values that describe them is why the engine can tell a connected body of pages is a complete treatment of a subject in the first place. That model is the mechanism of action behind this mechanism, and it has its own guide: entities, attributes, and why search ranks topics not keywords. This guide explains that complete coverage wins and why it beats budget; that guide explains the entity model that lets the engine see the coverage as authority, and it links back here for the reason a small site can exploit that model against a larger one. The boundary is clean: authority mechanism here, entity model there, and the two are reciprocal halves of the same answer rather than the same explanation told twice.

The four things that actually build topical authority

Four things build topical authority for a small business: breadth of coverage across the whole subject, internal connection that wires the pages into one body, depth on the few core pages that have to be the best on the web, and consistency of subject over time. These are not four tactics bolted onto a content plan. They are the four properties that, together, make an engine conclude a site is the source on a subject. A site can have one or two and not earn authority. It earns authority when it has all four on a subject narrow enough that having all four is achievable for a small operator. Each is treated below as its own thing, with what it is, why it builds authority, and what it looks like done by a real small business.

Breadth: covering the whole topic, not the head term

Breadth of coverage is treating the entire subject, the head question and every real sub-question a person who cares about it actually has, rather than a single page on the head term. It builds authority because the engine is asking whether a site covers the subject, and a subject is not one question. It is a connected set of them, and a site that resolves the whole set is demonstrably more of a source than a site that resolves one and stops.

For a regional commercial-roofing company, breadth on "commercial flat-roof replacement" is not one page repeating the phrase. It is the connected set: how to tell a repair from a replacement honestly, what each membrane type costs to own over its life, what a tear-off involves versus a recover, how weather and building use change the decision, what a warranty actually covers and what voids it, how a replacement is sequenced on an occupied building. Each is a real question a facilities manager has. A site that resolves all of them is, to the engine and to the reader, the source on commercial flat-roof replacement in a way a single head-term page can never be, regardless of how many links that page has. Without breadth there is no subject coverage to recognize, only a page, which is why it is the first builder and nothing else compensates for its absence.

Connection: the cluster has to be wired together

Connection is the internal linking that ties the pages on a subject into one navigable body so a machine and a reader can see the subject is covered as a whole, not as unrelated posts that happen to share a site. It builds authority because coverage that is not connected does not read as a body. Twenty disconnected pages on a subject look like twenty posts. The same twenty, wired so each reaches the related ones and the whole has a coherent shape, read as a complete treatment, which is the thing the engine is recognizing.

A specialty-parts supplier with fifteen unconnected articles on a component category is leaving most of the authority on the table even though the coverage exists, because nothing tells the engine or the buyer that these fifteen pages are one treatment of one subject. Wired so the spec-reading page reaches the substitution page, the substitution page reaches the failure-mode page, and the whole cluster has a clear internal shape, the same fifteen pages become a recognizable body and rank like one. This guide names connection as a builder and stops at the boundary: the actual architecture, link depth, anchor discipline, and cluster topology that does this well, is its own subject with its own guide, internal linking: the architecture that moves rankings. The point that belongs here is only that coverage without connection underperforms, and connection is the second of the four builders, not an optional polish step.

Depth on the core nodes: the few pages that have to be the best on the web

Depth on the core nodes is making the handful of pages at the center of the subject genuinely the best treatment of their question available anywhere, not adequate, the best. It builds authority because a body of coverage is anchored by its core pages, and if the pages everyone in the subject inevitably lands on are merely fine, the whole cluster is judged as merely fine. The breadth gives the engine a subject to recognize. The depth on the core nodes is what makes recognizing it worth the engine's while.

For the filtration-media site, the core nodes were the few pages every buyer hits: how to read the spec, how to validate a substitution, what each contaminant class requires. Those were not adequate. They were the most complete, most correct, most genuinely useful treatment of those exact questions on the web, written by someone who clearly did the work. The outer pages broadened the coverage; the core pages earned the authority, because they were the ones that made a procurement engineer and an engine both conclude this site is the source. A small business cannot make two hundred pages the best on the web. It can make the five or six that sit at the center of a narrow subject genuinely the best, which is the achievable, decisive form of depth: not depth everywhere, which is impossible, but depth where every searcher and the engine inevitably land.

Consistency of subject: why staying on one topic compounds and drifting dilutes

Consistency of subject is keeping the site's content on the defined subject over time so the signal compounds, rather than drifting across unrelated topics so the signal dilutes. It builds authority because the engine's judgment of what a site is the source on is formed over the whole site and over time, and every page off the subject is a page telling the engine the site is about something else. Coverage of one subject compounds into authority. Coverage spread thin across five unrelated subjects compounds into nothing in particular.

A regional landscaping company that publishes deeply on commercial grounds maintenance for one client type, then drifts into residential design, then into holiday lighting, then into snow removal, has produced volume and no authority, because no subject ever accumulated enough coherent coverage for the engine to conclude the site is the source on it. The same effort spent staying on commercial grounds maintenance, the head question and every real sub-question, compounds, because every page reinforces the same subject signal instead of starting a new one. For a small business with finite output, consistency is not a style preference. It is the difference between a body of work that becomes authority and a scattered archive that never does, and it is the builder most often lost not by choosing the wrong subject but by never committing to one at all.

Two things do not build it, and they are worth naming as plainly as the four that do, because they are the levers a small business is most often sold. A bigger backlink count is volume of an indirect, increasingly down-weighted proxy; it does not make a site the source on a subject the engine already has better evidence to judge. A bought domain rating is a vendor's number, not the engine's verdict; raising it raises a dashboard, not the rankings on the queries that pay. For the job of earning authority on a subject a small business can win, these are not weak builders to use sparingly. They are not builders, and the budget they consume is budget not spent on the four that are.

How to judge whether your own site has any authority yet

You can judge whether your site has earned any topical authority with one honest test: is this a genuinely complete, connected source on a defined subject, or is it a few pages and some links. The test is not a tool reading or a domain score. It is an honest read of whether the site, on a subject it could plausibly own, has the four builders or only the appearance of activity. Most small businesses, run honestly through this, find they have neither real coverage nor real authority yet, and that is useful, because it points the budget at the thing that would actually build it instead of the thing that has not.

The discipline is to judge the site against the mechanism, not against a vendor's number. A higher tool-reported domain rating with thin, disconnected, drifting content is not authority and will not rank on the queries that pay. A modest domain with complete, connected, deep, consistent coverage of a narrow subject is authority and will. The test below sorts which one a site actually is, and it takes an afternoon and no analytics stack.

The honest test: is this a complete source or a few pages and some links

Run each subject the business could own through four questions, honestly. Does the site cover the whole subject, the head question and the real sub-questions a buyer actually has, or one page on the head term and nothing around it. Are those pages connected into one navigable body, or are they unrelated posts that happen to share a domain. Are the few core pages at the center genuinely the best treatment of their question on the web, or merely adequate. Has the content stayed on this subject long enough to compound, or has it drifted across unrelated topics. Four yeses on a subject is earned authority on that subject. Mostly noes is a few pages and some links, no matter what a link tool reports.

A regional accounting firm ran this on "outsourced controller services for [an industry]" and found the honest answer was no on all four: one thin head-term page, nothing connected to it, the core page no better than three competitors', and a blog that wandered across unrelated tax-season topics. The link tool said the domain was respectable. The mechanism said there was no authority on that subject yet, and the mechanism was right, because the firm did not rank on the queries that produced engagements. The value of the test is that it gave an honest, actionable answer, build the coverage, where the dashboard had given a flattering, useless one.

What "complete enough to start ranking" looks like for a small niche

Complete enough to start ranking, for a narrow subject a small business can own, is not exhaustive coverage of an entire industry. It is the head question plus the genuine sub-questions a real buyer has, connected into one body, with the few core pages genuinely the best on the web for their exact question, held consistently on the subject. For a tightly defined niche that is frequently a focused cluster of pages, each resolving one real question, wired together, anchored by a handful of genuinely best-in-class core pages. Not hundreds. Enough to be unmistakably the complete source on a subject narrow enough to be ownable.

The judgment that matters is the scope of the subject, not a page count. The filtration-media site was complete enough at around thirty pages because the subject was defined that tightly: one media category, a specific set of contaminants. The same thirty pages spread across all industrial filtration would not have been complete enough for anything, because the subject would have been too broad for thirty pages to cover. Complete enough is a relationship between the size of the coverage and the narrowness of the subject. A small business wins by narrowing the subject until its achievable coverage is genuinely complete for that subject, not by trying to cover a broad subject thinly.

The first move: build the cluster, not the link pile

The first move for a small business with no authority yet is to commit to one defined subject it can plausibly own and start building the connected cluster that covers it, not to commission a link campaign. Pick the subject tightly enough that you can actually finish covering it. Identify the head question and the real sub-questions a buyer actually has. Commit to the few core pages being genuinely the best on the web for their question. Wire the pages together as they are built so the body is connected from the start. That is the move that earns authority, and it is the move a link budget cannot substitute for.

This guide states the first move and stops at the boundary. The disciplined procedure for decomposing a subject into the exact set of pages, sequencing them, and mapping the cluster before writing it, the topical map, is its own subject with its own guide: how to build a topical map for your business, which links back here for why building the cluster beats building the link pile. The point that belongs on this page is only the direction: the first move is the cluster, not the links, and the next page gives the map that turns "build the cluster" into a concrete plan.

Topical authority versus the things it gets confused with

Topical authority gets conflated with four near-neighbors, and each conflation sends a small business's budget at the wrong thing. Domain authority as a vendor score, backlink count, raw content volume, and the entity model that explains why engines cluster topics are each a different thing, and only one of them is what the engine actually ranks. Here is each one, what it actually is, and which guide owns it where this one stops.

Topical authority vs domain authority as a third-party score

Domain authority, or domain rating, is a third-party vendor's predictive score for how a domain might perform, computed largely from its link profile. Topical authority is the engine's actual judgment of whether a site is the source on a specific subject. The vendor score is a model's guess about a whole domain. Topical authority is the engine's verdict about a site on a subject. They are correlated enough that the score is not useless and different enough that acting on the score instead of the verdict is a mistake. A site can have a modest vendor score and dominate a defined subject because it is genuinely the source on it; a site can have a high vendor score and lose that subject because a higher number is not the same thing as being the source. Treat the vendor score as a rough external estimate, never as the thing to optimize, because the engine ranks the verdict, not the estimate of it.

Topical authority vs backlink count

Backlink count is the number of external links to a site. Topical authority is the completeness and coherence of a site's coverage of a subject. A count is not coverage. The link brokers sell the count because it is the thing they can produce, and it is the wrong lever for the queries a small business can win, where the engine has direct evidence of who the complete source is and leans far less on a link tally. The honest correction for a small business is not "links do not matter at all". It is "the count is not the thing that decides a defined subject you can own, and the budget aimed at growing the count is aimed away from the coverage that actually decides it". A small site that fixates on closing a link gap it cannot close ignores the contest it can win.

Topical authority vs raw content volume

Raw content volume is the number of pages or posts a site has published. Topical authority is whether those pages completely and coherently cover one defined subject. More pages is not more authority. Unstructured volume across scattered topics actively dilutes authority, because every off-subject page tells the engine the site is about something else and no subject ever accumulates enough coherent coverage to be recognized. A site with a hundred unconnected posts across ten subjects has volume and no authority; a site with twenty-five connected pages on one subject has authority and less volume. The lever is not how much was published but whether what was published makes the site the complete connected source on a subject. Volume without breadth, connection, depth, and consistency is activity, not authority.

Topical authority vs the entity model that explains it

The entity model, entities and the attributes and values that describe them, is the explanation of why a search engine can recognize a connected body of pages as a complete treatment of a subject and rank it as authoritative. Topical authority is what a site earns when its coverage satisfies that model better than its competitors. This guide owns the authority mechanism: that complete connected coverage wins and why it beats budget. It does not own the entity model that explains why the engine can see the coverage as authority at all; that is the mechanism of action behind this mechanism, and it has its own guide. The boundary, stated once and held: the why the engine clusters and ranks topics and entities model lives in entities, attributes, and why search ranks topics not keywords, this guide is the why a small site can exploit that to outrank a bigger one half, and the two link to each other as reciprocal halves rather than repeating each other. Read that guide for the model; this one is the mechanism a small business uses the model to win.

What earning authority changes around your business

Earning topical authority is not a setting you switch on. It changes what your content has to be, it changes what holds the site together, and it changes how durable the result is against bigger competitors and algorithm updates. These second-order effects are where the practical consequences land for a small business, and they are the reason this guide is an on-ramp to the rest of the pillar rather than the end of the topic.

How it changes the content you have to produce, from posts to a connected cluster

The first thing earning authority changes is what counts as the work. Under the link model, "doing SEO" could mean acquiring links to whatever the site already was. Under the authority mechanism, the work is producing a connected cluster that completely covers a defined subject: the head question and every real sub-question, the core pages genuinely the best on the web for their question, the whole thing wired together and held consistently on the subject over time. That is sustained editorial and structural work, not a campaign with an end date, and it is a different and larger commitment than publishing occasional posts or buying occasional links.

For most small businesses this is the honest sticking point. The work is real, it is editorial and structural, it has to be done well and kept on subject over quarters, and it is not something a busy ten-to-two-hundred-person company typically has the in-house capacity to execute, because being the complete connected source on a subject is a job, not a side task. Doing that work, turning a site into the genuinely complete, connected, best-in-class source on a subject a business can own, is exactly what Iron Goo's SEO service exists to execute for companies that do not staff it internally. That is the contextually honest bridge: the mechanism is real and winnable, and the reason most SMBs do not win it is not that they cannot, it is that nobody on the team owns building the cluster.

How it changes what holds the site together

Earning authority changes the load-bearing element of the site from individual pages to the connections between them. Once authority is a property of a connected body rather than a stack of separate pages, the internal links that wire the body together stop being navigation and become structural: they are part of what makes the engine read the pages as one treatment of one subject. A cluster with the right coverage but the wrong connection underperforms the same cluster wired correctly, which is why connection is one of the four builders and not an afterthought.

This guide names that shift and stops at the boundary, because the actual architecture is its own subject. How deep the linking should go, how anchors should be written, how the cluster's topology should be shaped so authority concentrates where it should, is owned by internal linking: the architecture that moves rankings. The point that belongs here is only the second-order fact: earning authority changes what holds the site together from pages to the structured connections between them, and the architecture for doing that well is handed off, not taught here.

Why authority earned by coverage survives a bigger competitor and an update

The most valuable second-order effect is durability. Authority earned by being the genuinely complete, connected source on a subject is durable against both a bigger competitor and an algorithm update, for the same reason in both cases: it is built on the thing the engine is actually trying to reward, so a bigger competitor cannot out-link it on that subject and an update tuned to find the best source keeps finding it. A link-bought position is fragile because it is a bet against a specific proxy the engine is steadily de-weighting, and an update that re-weights it can erase the position with nothing on the site having changed. A coverage-earned position is aligned with what the engine keeps moving toward, so the same work that ranks it today is what the next update is trying to surface.

A national competitor with a hundred times the links cannot easily take a defined subject from a small site that genuinely owns it, because the competitor would have to actually become the more complete, more connected, deeper, more consistent source on that specific subject, which is the small specialist's whole focus and a small slice of the generalist's sprawling site. That asymmetry is the small business's structural advantage and it is exactly what makes coverage-earned authority durable in a way a bought position never is. The mechanism does not just let a small site win the subject. It lets a small site hold it.

Where this leaves you, and the first cluster node to commit to

Topical authority is how an SMB earns durable search visibility in an era where search itself increasingly answers the question and rewards the actual source over the biggest spender: not by out-buying links it cannot out-buy, but by out-covering a subject it can genuinely own, because on the queries a small business can win, out-covering the subject is what decides the result. That is the whole reason a small site beats a big one, and it is durable for the same reason it works, because it is built on the thing the engine keeps moving toward rather than the proxy it keeps moving away from.

This page explained the mechanism and held its edges on purpose, handing the entity model, the map procedure, the link architecture, and the snippet craft to the guides that own them rather than blurring them. The two that come next are the reciprocal half and the concrete plan: entities, attributes, and why search ranks topics not keywords is the model that explains why the engine can read your coverage as authority at all, and how to build a topical map for your business is what turns "build the cluster" into a sequenced set of pages. The most useful next action is not "get more links" and not "publish more". It is narrower: pick the one subject your business can plausibly be the genuine source on, name the single core page that has to be the best on the web for its exact question, commit to building that page first, and read the entity-model guide next so you understand why the coverage you are about to build is the thing the engine will read as authority.